Loading data

setwd("/data02/Analysis/Projects/2_CPE_Transmission/Kmer_Comparison_ForPlasmidClustering")

library(igraph)
library(ggplot2)
library(circlize)
library(dplyr)
library(ggrepel)
library(kableExtra)
library(InteractiveComplexHeatmap)
library(ComplexHeatmap)

plasmid_similarity_df <- data.table::fread("List_1071_Plasmid_seqs_kmer_comp_Jaccard_v2.tab", header= FALSE, sep = "\t")
#head(plasmid_similarity_df)

colnames(plasmid_similarity_df) <- c("Plasmid1","Plasmid1_length","Plasmid2","Plasmid2_length", "Plasmid_similarity","Plasmid_disimilarity")

Plasmid Clustering

# Generate a similarity matrix
  simMatrix  <- plasmid_similarity_df %>% 
    # head() %>%
    select(Plasmid1, Plasmid2, Plasmid_similarity) %>% 
    data.table::dcast(Plasmid1~Plasmid2, value.var="Plasmid_similarity") %>% 
    replace(is.na(.), 1)
  
  simMatrix <- simMatrix %>% tibble::column_to_rownames("Plasmid1")
  
  simMatrix[is.na(simMatrix)] <- 1
  
  p1 <- Heatmap(as.matrix(simMatrix), name = "Plasmid Similarity", show_row_names = FALSE, show_column_names = FALSE)
  
Closed_CP_Plasmids_metadata_df <- data.table::fread("/data02/Analysis/Projects/2_CPE_Transmission/Reference_Excels/Closed_CP_Plasmids_metadata_07102022.csv", header= TRUE, sep = ",")
  
  Closed_CP_Plasmids_metadata_subsetdf <- Closed_CP_Plasmids_metadata_df %>%  select(Plasmid_name_filename,CP_Gene,Hospital,Sample_type,Inc_group_regrouped)

Plas_CPdf <- Closed_CP_Plasmids_metadata_subsetdf %>%  select(Plasmid_name_filename,CP_Gene)
Plas_CPdf <- Plas_CPdf %>% tibble::column_to_rownames("Plasmid_name_filename")

p2 <- Heatmap(Plas_CPdf, name = "CP Genes", show_row_names = FALSE)

#p1

#ht_list <- p1 + p2

#ht_list 
#htShiny(ht_list)


  # col_fun = colorRamp2(c(0, 0.25, 0.5, 0.75, 1), c( "#06d6a0", "#80cfba", "#ffd166", "#eda4b5", "#ef476f"))
  # #col_fun
  # Heatmap(as.matrix(simMatrix), name = "Plasmid Similarity", show_row_names = FALSE, show_column_names = FALSE,col = col_fun)
 # Heatmap(as.matrix(simMatrix), name = "Plasmid Similarity", show_row_names = FALSE, show_column_names = FALSE,col = col_fun, row_km = 7, column_km = 5)
  
    col_fun = colorRamp2(c(0, 0.25, 0.5, 0.75, 0.9, 1), c("#0232de", "#4e6dde", "#ffd166", "#fabb28", "#ef476f", "#e30742")) 
  
  Heatmap(as.matrix(simMatrix), name = "Plasmid Similarity", show_row_names = FALSE, show_column_names = FALSE,col = col_fun,
  heatmap_legend_param = list(
    title = "Plasmid Similarity", at = c(0, 0.25, 0.5, 0.75, 0.9, 1), 
    labels = c(0, 0.25, 0.5, 0.75, 0.9, 1)
  ))

Cumulative plots

Histogram1

Plasmid Similarity histogram (Bin size 0.02)

# plasmid_similarity_df %>% 
#   ggplot(aes(x=Plasmid_similarity)) +
#   geom_histogram(binwidth=0.02,color="black",fill = "#4dbfbc") +
#   scale_x_continuous(breaks = seq(0, 20, 1)) +
#   stat_bin(binwidth=0.02, geom="text", aes(label=..count..), vjust=-1.5)+
#   labs(x = "Plasmid Similarity (Jaccard)", y = "Plasmid Pair Counts") +
#   ggtitle("Histogram of Plasmid Similarity (Bin size 0.02)") +
#   theme_classic()

#NOTE: The above geom_histogram did not give me expected counts. So, I calculated ranges first and then tabled and then made the dataframe which was then supplied to ggplot to draw a plot

stack(table(cut(plasmid_similarity_df$Plasmid_similarity,breaks=seq.int(from=0,to=1,by=0.02)))) %>% 
  as.data.frame() %>% 
  rename(PlasmidPairCount=values, Range=ind) %>% 
  ggplot(aes(x=Range,y=PlasmidPairCount, label = PlasmidPairCount)) + 
  geom_col(color="black",fill = "#4dbfbc",position = "dodge") +
   geom_text(position = position_dodge(width = 1), hjust = -0.1) +
  labs(x = "Plasmid Similarity (Jaccard)", y = "Plasmid Pair Counts") +
  ggtitle("Histogram of Plasmid Similarity (Bin size 0.02)") +
  theme_classic() +
    theme(axis.text.y=element_text(size=10,  vjust = 0, hjust = 1),
        axis.text.x=element_text(size=10, angle=90, vjust = 0, hjust = 1.2),
        axis.line=element_blank(),
        axis.ticks=element_blank(),
        legend.position="top") +
  coord_flip()

Histogram2

Plasmid Similarity histogram (Bin size 0.05)

# plasmid_similarity_df %>% 
#   ggplot(aes(x=Plasmid_similarity)) +
#   geom_histogram(binwidth=0.05,color="black",fill = "#4dbfbc") +
#   #scale_x_continuous(breaks = seq(0, 20, 1), limits = c(0, 1)) + 
#   stat_bin(binwidth=0.05, geom="text", aes(label=..count..), vjust=-1.5)+
#   labs(x = "Plasmid Similarity (Jaccard)", y = "Plasmid Pair Counts") +
#   ggtitle("Histogram of Plasmid Similarity (Bin size 0.05)") +
#   theme_classic()
# 
# plasmid_similarity_df %>% 
#   ggplot(aes(x=Plasmid_similarity)) +
# geom_histogram(
#   color="black",fill = "#4dbfbc",
#     breaks = seq(0, 1, by = 0.05), 
#     aes(fill = ..count..), position = "identity") + 
#   #scale_y_log10() +
#     scale_x_continuous(breaks = seq(0, 1, by=0.05)) +
#   stat_bin(binwidth=0.05, geom="text", aes(label=..count..), vjust=-1.5, hjust=-0.2)+
#     labs(x = "Plasmid Similarity (Jaccard)", y = "Plasmid Pair Counts") +
#   ggtitle("Histogram of Plasmid Similarity (Bin size 0.05)") +
#   theme_classic()
# 
# plasmid_similarity_df %>% 
#   ggplot(aes(x=Plasmid_similarity)) +
#     stat_bin(binwidth=0.02) +  
#     stat_bin(binwidth=0.02, geom="text", aes(label=..count..), vjust=-1.5) +
#   scale_x_continuous(breaks = seq(0, 1, by=0.02)) 

stack(table(cut(plasmid_similarity_df$Plasmid_similarity,breaks=seq.int(from=0,to=1,by=0.05)))) %>% 
  as.data.frame() %>% 
  rename(PlasmidPairCount=values, Range=ind) %>% 
  ggplot(aes(x=Range,y=PlasmidPairCount, label = PlasmidPairCount)) + 
  geom_col(color="black",fill = "#4dbfbc",position = "dodge") +
   geom_text(position = position_dodge(width = 1), hjust = -0.1) +
  labs(x = "Plasmid Similarity (Jaccard)", y = "Plasmid Pair Counts") +
  ggtitle("Histogram of Plasmid Similarity (Bin size 0.05)") +
  theme_classic() +
    theme(axis.text.y=element_text(size=10,  vjust = 0, hjust = 1),
        axis.text.x=element_text(size=10, angle=90, vjust = 0, hjust = 1.2),
        axis.line=element_blank(),
        axis.ticks=element_blank(),
        legend.position="top") +
  coord_flip()

Full

Cumulative frequency plot

plasmid_similarity_df %>% 
  count(Plasmid_similarity) %>% 
  mutate(cumsum_n = cumsum(n)) %>% 
  ggplot(aes(x=Plasmid_similarity, y=cumsum_n)) +
  geom_point(color="red", size=0.1) +
  theme_classic()

Cutoff 0.97

Cumulative frequency plot at plasmid similarity 0.97 cutoff

# Cumulative frequency plot from 0.9
plasmid_similarity_df %>% 
  count(Plasmid_similarity) %>% 
  mutate(cumsum_n = cumsum(n)) %>%
  filter(Plasmid_similarity>=0.97) %>% 
  ggplot(aes(x=Plasmid_similarity, y=cumsum_n)) +
  geom_point(color="red", size=0.1) +
  theme_classic()

Cutoff 0.9

Cumulative frequency plot at plasmid similarity 0.9 cutoff

# Cumulative frequency plot from 0.9
plasmid_similarity_df %>% 
  count(Plasmid_similarity) %>% 
  mutate(cumsum_n = cumsum(n)) %>%
  filter(Plasmid_similarity>=0.9) %>% 
  ggplot(aes(x=Plasmid_similarity, y=cumsum_n)) +
  geom_point(color="red", size=0.1) +
  theme_classic()

Cutoff 0.8

Cumulative frequency plot at plasmid similarity 0.9 cutoff

# Cumulative frequency plot from 0.9
plasmid_similarity_df %>% 
  count(Plasmid_similarity) %>% 
  mutate(cumsum_n = cumsum(n)) %>%
  filter(Plasmid_similarity>=0.8) %>% 
  ggplot(aes(x=Plasmid_similarity, y=cumsum_n)) +
  geom_point(color="red", size=0.1) +
  theme_classic()

#summary(plasmid_similarity_df)

for(sim_cutoff in seq(0.80, 1, by = 0.01)) {

Plaus_ClustrPairs <- plasmid_similarity_df %>% 
  filter(Plasmid_similarity >= sim_cutoff) %>% 
  select(Plasmid1,Plasmid2)
  nrow(Plaus_ClustrPairs)

 
  g1 <- graph.data.frame(Plaus_ClustrPairs, directed = FALSE)
  g1
  #plot(g1)
  
  #Fancy figure start
  #n = 100
  #p = 1.5/n
  #g = erdos.renyi.game(n, p)
  #coords = layout.fruchterman.reingold(g)
  #plot(g1, layout=coords, vertex.size = 3, vertex.label=NA)
  #plot(g1, vertex.label=NA, vertex.size=2, vertex.color="#0CCF02")
  #Fancy figure end
  
  
  # To generate Sample and their ClusterNumber
  cl1 <- clusters(g1)
  
  tbl1 <- cbind( V(g1)$name, cl1$membership )
  #class(tbl1)
  tbl1 <- as.data.frame(tbl1)
  colnames(tbl1) = c("sample", "Cluster")
  tbl1 <- tbl1 %>% arrange(as.numeric(Cluster)) 
  
  write.table(tbl1,file=paste0("Samples_ClusterNumber_sim_cutoff",sim_cutoff,".txt"),row.names=FALSE) # drops the rownames and write to text file
}  
awk '{print FILENAME "\t" $0}' Samples*.txt | sed -e 's/Samples_ClusterNumber_sim_cutoff//g' -e 's/.txt//g' | grep -v 'sample' | sed 's/"//g' | awk '{print $1 "\t" $2 "\t" $3}' >Combined_cutoff_0.8to1_sample_clusternumber.tab

Plasmid and Cluster Count plots

Table

Similarity Cutoff, Plasmid Counts and ClusterCounts for each cutoff

cutoff_sample_clusternumber_df <- data.table::fread("Combined_cutoff_0.8to1_sample_clusternumber.tab", header= FALSE, sep = "\t")
#head(cutoff_sample_clusternumber_df)

colnames(cutoff_sample_clusternumber_df) <- c("Cutoff","Plasmid","ClusterNumber")

Cutoff_PlasmidCount <- cutoff_sample_clusternumber_df %>% 
  count(Cutoff) %>% 
  as.data.frame() %>% 
  rename(PlasmidCount=n) 

Cutoff_ClusterCount <-cutoff_sample_clusternumber_df %>% 
  group_by(Cutoff) %>% 
  summarise(max = max(ClusterNumber, na.rm=TRUE)) %>% 
  as.data.frame() %>% 
  rename(ClusterCount=max) 

Cutoff_PlasmidCount_ClusterCount <- left_join(Cutoff_PlasmidCount, Cutoff_ClusterCount, by="Cutoff")

Cutoff_PlasmidCount_ClusterCount %>% 
  kbl(caption = "Similarity Cutoff, Plasmid Counts and ClusterCounts for each cutoff") %>%
  kable_classic(full_width = F, html_font = "Cambria")
Similarity Cutoff, Plasmid Counts and ClusterCounts for each cutoff
Cutoff PlasmidCount ClusterCount
0.80 1010 40
0.81 1010 40
0.82 1009 40
0.83 1009 41
0.84 1009 41
0.85 1009 41
0.86 1006 40
0.87 1005 41
0.88 1005 42
0.89 1004 42
0.90 1001 43
0.91 999 43
0.92 997 45
0.93 995 44
0.94 990 46
0.95 977 47
0.96 962 45
0.97 955 50
0.98 949 52
0.99 931 58
1.00 468 74

Plot1

Similarity Cutoff, Plasmid Counts and ClusterCounts for each cutoff

Cutoff_PlasmidCount_ClusterCount %>% 
  ggplot(aes(PlasmidCount,ClusterCount), color=Cutoff) +
  geom_point(color="blue", size=1) +
    geom_label_repel(aes(label = Cutoff),
                  box.padding   = 0.35, 
                  point.padding = 0.5,
                  segment.color = 'grey50') +
  theme_classic()

Plot2

Similarity Cutoff, Plasmid Counts and ClusterCounts for each cutoff

Cutoff_PlasmidCount_ClusterCount %>% 
  filter(Cutoff!=1) %>% 
  ggplot(aes(PlasmidCount,ClusterCount), color=Cutoff) +
  geom_point(color="blue", size=1) +
  scale_x_continuous(breaks=seq(920, 1020, 10)) +
    geom_label_repel(aes(label = Cutoff),
                  box.padding   = 0.35, 
                  point.padding = 0.5,
                  segment.color = 'grey50') +
  theme_classic()

Plasmid Clusters information table

CP_Plasmids_meta_ClusterNumber_cutoff0.9_df <- data.table::fread("Closed_CP_Plasmids_metadata_07102022.tab", header= TRUE, sep = "\t")
#head(CP_Plasmids_meta_ClusterNumber_cutoff0.9_df)

Clusters_info_df <- CP_Plasmids_meta_ClusterNumber_cutoff0.9_df %>% 
  select(ClusterNumber_based_on_0.9_cutoff, Inc_group_regrouped, Relaxase_type_corrected, Length,ENT_ID,PID) %>% 
  group_by(ClusterNumber_based_on_0.9_cutoff) %>% 
  summarise(Inc_group_regrouped_collapse=paste(unique(Inc_group_regrouped),collapse='#'),
            Relaxase_type_corrected_collapse=paste(unique(Relaxase_type_corrected),collapse='#'),
            IsolateCount=n(),
            PatientCount = n_distinct(PID),
            PlasmidLength_Min = min(Length),
            PlasmidLength_Max = max(Length),
            ) %>% 
  rename(ClusterNum=ClusterNumber_based_on_0.9_cutoff,
         Inc_Group=Inc_group_regrouped_collapse,
         Relaxase=Relaxase_type_corrected_collapse) 

Clusters_info_df %>% 
  kbl(caption = "Plasmid Clusters information table") %>% 
  kable_classic(full_width = F, html_font = "Cambria") %>% 
  column_spec(1, width = "1px")
Plasmid Clusters information table
ClusterNum Inc_Group Relaxase IsolateCount PatientCount PlasmidLength_Min PlasmidLength_Max
1 IncU
7 7 7580 7580
10 IncH MOBH 3 3 306281 315533
11 IncR
12 10 74789 78200
12 IncN MOBF 2 1 59471 59567
13 IncF MOBP 4 4 135212 135220
14 IncL/M MOBP 4 3 67445 71114
15 IncF MOBF 4 4 173234 180817
16 IncA/C2 MOBH,MOBH 34 27 174317 176865
17 IncN MOBF 4 4 49992 50752
18 IncH MOBH 3 3 310749 312015
19 IncL/M MOBP 33 23 63589 69945
2 IncU#ColRNAI_all MOBP 405 203 67304 80977
20 IncF MOBC 3 2 108331 108331
21 IncA/C2_IncF MOBH,MOBH 4 1 173887 173888
22 IncH MOBH 2 2 278408 291650
23 IncA/C2 MOBH,MOBH 2 2 146179 149464
24 IncX3 MOBP 7 6 51479 51503
25 IncF MOBF 8 7 103933 112627
26 rep_cluster_1195 MOBP 15 11 6139 12279
27 IncF MOBF 2 2 152661 161498
28 IncF MOBF 3 3 118971 122714
29 IncF MOBF 2 2 113639 113641
3 IncU
2 1 21549 21549
30 ColRNAI_all MOBF,MOBP 2 1 142050 142353
31 IncA/C2 MOBH,MOBH 4 4 160819 164596
32 IncU MOBP 25 16 78125 99518
33 IncF_combinations MOBF 2 1 128051 128105
34 IncF_IncU_IncX MOBF,MOBF 7 5 121423 123974
35 IncN_IncU MOBF 8 6 70461 76006
36 IncF_combinations MOBF 8 5 111676 112911
37 IncL/M
4 4 50138 50138
38 IncF_combinations MOBF 2 2 177751 198148
39 IncA/C2
2 1 88057 88057
4 IncF_combinations MOBF 2 2 105819 105819
40 IncN_IncU_IncX MOBP 3 1 97507 97513
41 IncF MOBF 2 2 30320 60013
42 IncF MOBF 4 3 63637 93113
43 IncH MOBH 4 3 266420 300349
5 IncL/M MOBP 28 21 82120 91643
6 IncN MOBF 286 211 38383 45284
7 IncL/M MOBP 6 6 82759 91289
8 IncA/C2 MOBH,MOBH 19 12 117950 127481
9 IncX3 MOBP 18 16 45048 46161
Singleton IncF_IncU#IncF#IncX3#IncA/C2_IncF#IncH#IncA/C2#rep_cluster_1195#IncF_combinations#IncN#ColRNAI_all#IncN_IncU#IncU_IncX#IncU#IncA/C2_IncN#IncN_IncU_IncX#IncI_IncU#IncR MOBF#-#MOBF,MOBH,MOBH#MOBH,MOBH#MOBH#MOBP#MOBF,MOBF#MOBP,MOBP#MOBF,MOBH#MOBC 70 62 6494 476834 
#write.table(Clusters_info_df,file="Clusters_info_datatable.txt",sep = "\t", row.names=FALSE) # drops the rownames and write to text file

Plasmid persistence

We are trying to check how the plasmid is persistent across the year whereas the clonal cluster outbreaks are sporadic and short-lived.

library(lubridate)
library(ggplot2)
library(forcats)
library(glue)

meta <- read.table("/data02/Analysis/Projects/2_CPE_Transmission/Plasmid_roary/Plasmids_1071_PCA_plot/Closed_CP_Plasmids_1071_complete_v3_Plasmid_meta_for_gubbins.csv", sep = ",", header = TRUE)

Phylodynamics_staticplot <- function(plot_title,clusterNum){
  #View(meta)
  # Subsetting to a smaller set for ease of simplicity - selecting OXA-48 Plasmids records with n=41 rows
 subset_clusterNum <- Closed_CP_Plasmids_metadata_df %>% 
  filter(ClusterNumber_based_on_0.9_cutoff==clusterNum) %>%  # Although the variable number
  mutate(Unique_Plasmid_Cluster = if_else(is.na(Clonal_cluster),paste0("Singleton_", row_number()), as.character(Clonal_cluster))) %>%  # assign unique id for NA values
    group_by(Unique_Plasmid_Cluster) %>% 
    mutate(group_id = cur_group_id())
  
subset_clusterNum$Date_of_culture <- as.Date(subset_clusterNum$Date_of_culture, format = "%d/%m/%Y")
  
#View(subset_clusterNum)
  
pd1 <- subset_clusterNum %>%
  ungroup() %>% 
  mutate(Unique_Plasmid_Cluster = forcats::fct_reorder(Unique_Plasmid_Cluster, Date_of_culture, min)) %>%
  ggplot(aes(Date_of_culture, Unique_Plasmid_Cluster,color = Lab_Species, shape= factor(Hospital))) +
  geom_line(aes(group=Unique_Plasmid_Cluster), color='black') +
  geom_point(size=3) +
  ggtitle(plot_title) +
  ylab("Cluster") + xlab("Date") +
  theme_bw() +
  theme(
    panel.grid.major = element_blank(),
    legend.position = "right",
    legend.key=element_rect(fill='gray96'),
    legend.title =element_text(size=10),
    text=element_text(size=12),
    axis.title.x = element_text(vjust = 0, size = 11),
    axis.title.y = element_text(vjust = 2, size = 11),
    axis.text.x = element_text(angle = 90, hjust = 1, size = 9),
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank())
pd1

}

Phylodynamics_dynamicplot <- function(plot_title,clusterNum){
    #View(meta)
  # Subsetting to a smaller set for ease of simplicity - selecting OXA-48 Plasmids records with n=41 rows
 subset_clusterNum <- Closed_CP_Plasmids_metadata_df %>% 
  filter(ClusterNumber_based_on_0.9_cutoff==clusterNum) %>%  # Although the variable number
  mutate(Unique_Plasmid_Cluster = if_else(is.na(Clonal_cluster),paste0("Singleton_", row_number()), as.character(Clonal_cluster))) %>%  # assign unique id for NA values
    group_by(Unique_Plasmid_Cluster) %>% 
    mutate(group_id = cur_group_id())
  
subset_clusterNum$Date_of_culture <- as.Date(subset_clusterNum$Date_of_culture, format = "%d/%m/%Y")
  
#View(subset_clusterNum)
  
pd1 <- subset_clusterNum %>%
  ungroup() %>% 
  mutate(Unique_Plasmid_Cluster = forcats::fct_reorder(Unique_Plasmid_Cluster, Date_of_culture, min)) %>%
  ggplot(aes(Date_of_culture, Unique_Plasmid_Cluster,color = Lab_Species, shape= factor(Hospital))) +
  geom_line(aes(group=Unique_Plasmid_Cluster), color='black') +
  geom_point(size=3) +
  ggtitle(plot_title) +
  ylab("Cluster") + xlab("Date") +
  theme_bw() +
  theme(
    panel.grid.major = element_blank(),
    legend.position = "right",
    legend.key=element_rect(fill='gray96'),
    legend.title =element_text(size=10),
    text=element_text(size=12),
    axis.title.x = element_text(vjust = 0, size = 11),
    axis.title.y = element_text(vjust = 2, size = 11),
    axis.text.x = element_text(angle = 90, hjust = 1, size = 9),
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank())
#pd1

pd1_plotly <- plotly::ggplotly(pd1)
pd1_plotly
}

Plasmid persistence plots (Static)

blaNDM-1

Phylodynamics_staticplot(plot_title="Phylodynamics of blaNDM-1 gene tranmission",clusterNum=6)

blaKPC-2

Phylodynamics_staticplot(plot_title="Phylodynamics of blaKPC-2 gene tranmission",clusterNum=2)

blaOXA-181

Phylodynamics_staticplot(plot_title="Phylodynamics of blaOXA-181 gene tranmission",clusterNum=16)

blaOXA-48

Phylodynamics_staticplot(plot_title="Phylodynamics of blaOXA-48 gene tranmission",clusterNum=19)

blaNDM-1

Phylodynamics_staticplot(plot_title="Phylodynamics of blaNDM-1 gene tranmission",clusterNum=5)

Plasmid persistence plots (Dynamic)

blaNDM-1

Phylodynamics_dynamicplot(plot_title="Phylodynamics of blaNDM-1 gene tranmission",clusterNum=6)

blaKPC-2

Phylodynamics_dynamicplot(plot_title="Phylodynamics of blaKPC-2 gene tranmission",clusterNum=2)

blaOXA-181

Phylodynamics_dynamicplot(plot_title="Phylodynamics of blaOXA-181 gene tranmission",clusterNum=16)

blaOXA-48

Phylodynamics_dynamicplot(plot_title="Phylodynamics of blaOXA-48 gene tranmission",clusterNum=19)

blaNDM-1

Phylodynamics_dynamicplot(plot_title="Phylodynamics of blaNDM-1 gene tranmission",clusterNum=5)